Generalized methods for algorithm development on optical systems

A. Al-Ayyoub, A. Awwad, K. Day, M. Ould-Khaoua

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

A number of recent studies have revealed that the Optical Transpose Interconnection Systems (or OTIS) are promising candidates for future high-performance parallel computers. In this paper, we present and evaluate two general methods for algorithm development on the OTIS. The proposed methods are general in the sense that no specific factor network or problem domain is assumed. The proposed methods allow efficient mapping of a wide class of algorithms into the OTIS. These methods are based on grids and pipelines as popular structures that support a vast body of parallel applications including linear algebra, divide-and-conquer type of algorithms, sorting, and FFT computation. Timing models for measuring the performance of the proposed methods are also provided. Through these models, the performance of various algorithms on the OTIS are evaluated and compared with their counterparts on conventional electronic interconnection systems. This study confirms the viability of the OTIS as an attractive alternative for large-scale parallel architectures. Finally, we show how the proposed methods can be used to design parallel algorithms for linear algebra on the OTIS.

Original languageEnglish
Pages (from-to)111-125
Number of pages15
JournalJournal of Supercomputing
Volume38
Issue number2
DOIs
Publication statusPublished - Nov 2006

Fingerprint

Optical systems
Optical System
Linear algebra
Interconnection
Optical interconnects
Parallel architectures
Sorting
Parallel algorithms
Fast Fourier transforms
Transpose
Sorting algorithm
Pipelines
Divide and conquer
Parallel Architectures
Parallel Applications
Parallel Computers
Viability
Parallel Algorithms
Timing
High Performance

Keywords

  • Interconnection networks
  • Optoelectronic systems
  • OTIS
  • Parallel algorithms

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Hardware and Architecture
  • Theoretical Computer Science
  • Computational Theory and Mathematics

Cite this

Generalized methods for algorithm development on optical systems. / Al-Ayyoub, A.; Awwad, A.; Day, K.; Ould-Khaoua, M.

In: Journal of Supercomputing, Vol. 38, No. 2, 11.2006, p. 111-125.

Research output: Contribution to journalArticle

@article{0fc7f0b56d094a07958c6118087054ee,
title = "Generalized methods for algorithm development on optical systems",
abstract = "A number of recent studies have revealed that the Optical Transpose Interconnection Systems (or OTIS) are promising candidates for future high-performance parallel computers. In this paper, we present and evaluate two general methods for algorithm development on the OTIS. The proposed methods are general in the sense that no specific factor network or problem domain is assumed. The proposed methods allow efficient mapping of a wide class of algorithms into the OTIS. These methods are based on grids and pipelines as popular structures that support a vast body of parallel applications including linear algebra, divide-and-conquer type of algorithms, sorting, and FFT computation. Timing models for measuring the performance of the proposed methods are also provided. Through these models, the performance of various algorithms on the OTIS are evaluated and compared with their counterparts on conventional electronic interconnection systems. This study confirms the viability of the OTIS as an attractive alternative for large-scale parallel architectures. Finally, we show how the proposed methods can be used to design parallel algorithms for linear algebra on the OTIS.",
keywords = "Interconnection networks, Optoelectronic systems, OTIS, Parallel algorithms",
author = "A. Al-Ayyoub and A. Awwad and K. Day and M. Ould-Khaoua",
year = "2006",
month = "11",
doi = "10.1007/s11227-006-7447-6",
language = "English",
volume = "38",
pages = "111--125",
journal = "Journal of Supercomputing",
issn = "0920-8542",
publisher = "Springer Netherlands",
number = "2",

}

TY - JOUR

T1 - Generalized methods for algorithm development on optical systems

AU - Al-Ayyoub, A.

AU - Awwad, A.

AU - Day, K.

AU - Ould-Khaoua, M.

PY - 2006/11

Y1 - 2006/11

N2 - A number of recent studies have revealed that the Optical Transpose Interconnection Systems (or OTIS) are promising candidates for future high-performance parallel computers. In this paper, we present and evaluate two general methods for algorithm development on the OTIS. The proposed methods are general in the sense that no specific factor network or problem domain is assumed. The proposed methods allow efficient mapping of a wide class of algorithms into the OTIS. These methods are based on grids and pipelines as popular structures that support a vast body of parallel applications including linear algebra, divide-and-conquer type of algorithms, sorting, and FFT computation. Timing models for measuring the performance of the proposed methods are also provided. Through these models, the performance of various algorithms on the OTIS are evaluated and compared with their counterparts on conventional electronic interconnection systems. This study confirms the viability of the OTIS as an attractive alternative for large-scale parallel architectures. Finally, we show how the proposed methods can be used to design parallel algorithms for linear algebra on the OTIS.

AB - A number of recent studies have revealed that the Optical Transpose Interconnection Systems (or OTIS) are promising candidates for future high-performance parallel computers. In this paper, we present and evaluate two general methods for algorithm development on the OTIS. The proposed methods are general in the sense that no specific factor network or problem domain is assumed. The proposed methods allow efficient mapping of a wide class of algorithms into the OTIS. These methods are based on grids and pipelines as popular structures that support a vast body of parallel applications including linear algebra, divide-and-conquer type of algorithms, sorting, and FFT computation. Timing models for measuring the performance of the proposed methods are also provided. Through these models, the performance of various algorithms on the OTIS are evaluated and compared with their counterparts on conventional electronic interconnection systems. This study confirms the viability of the OTIS as an attractive alternative for large-scale parallel architectures. Finally, we show how the proposed methods can be used to design parallel algorithms for linear algebra on the OTIS.

KW - Interconnection networks

KW - Optoelectronic systems

KW - OTIS

KW - Parallel algorithms

UR - http://www.scopus.com/inward/record.url?scp=33748700798&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33748700798&partnerID=8YFLogxK

U2 - 10.1007/s11227-006-7447-6

DO - 10.1007/s11227-006-7447-6

M3 - Article

VL - 38

SP - 111

EP - 125

JO - Journal of Supercomputing

JF - Journal of Supercomputing

SN - 0920-8542

IS - 2

ER -